Physical protection of soil carbon in macroaggregates does not reduce the temperature dependence of soil CO2 emissions
Résumé
In a warmer world, soil CO2 emissions are likely to increase. There is still much discussion about which soil organic C (SOC) pools are more sensitive to increasing temperatures. While the temperature sensitivity of C stabilized by biochemical recalcitrance or by sorption to mineral surfaces has been characterized, few studies have been carried out on the temperature sensitivity —expressed as Q10— of C physically protected inside soil macroaggregates (0.2–2 mm). It has been suggested that increasing the availability of labile SOC by exposing C through macroaggregate crushing would decrease Q10, i.e., the temperature dependence of soil CO2 emissions.
To test this hypothesis, the temperature dependence of CO2 emissions from C physically protected in macroaggregates was measured through 21-d laboratory incubations of crushed and uncrushed soils, at 18°C and 28°C. 199 topsoil samples, acidic or calcareous, with SOC ranging from 2 to 121 g kg−1 soil were investigated.
The CO2 emissions were slightly more sensible to temperature than to C deprotection: about 0.3 mg C g−1 soil (= 13 mg C g−1 SOC) and 0.2 mg C g−1 (= 12 mg C g−1 SOC) were additionally mineralized, in average, by increasing the temperature or by disrupting the soil structure, respectively. The mean Q10 index ratio of CO2 emitted at 28°C and 18°C was similar for crushed and uncrushed soil samples and equaled 1.6. This was partly explained because Q10 of macro-aggregate-protected C was 1. The results did not support the initial hypothesis of lower temperature dependence of soil CO2 emissions after macroaggregate disruption, although a slight decrease of Q10 was noticeable after crushing for soils with high amounts of macroaggregate-protected C. Field research is now needed to confirm that soil tillage might have no effect on the temperature sensitivity of SOC stocks.
Origine | Fichiers produits par l'(les) auteur(s) |
---|---|
Licence |